The broad long term objective is to determine the mechanism and role in development of the epigenetic -system of genomic imprinting. Imprinting of the H19 and insulin-like growth factor 2 (Igf2) genes (resulting in their monoallelic expression) is achieved through paternal-specific methylation of a 2.4 kb imprinting control region (ICR) as inherited from the male germ line. The specific aims are designed to further understanding of (1) how male germ cell-specific methylation of the ICR is achieved-this would be the imprinting mechanism, and (2) how the maternal ICR functions as a chromatin insulator in somatic cells. The specific aims are (1) in somatic cells, to map the ICR for sites of parent-of-origin allele-specific protein binding, and similarly, in germ cells, to map the ICR for sites of sex-specific protein binding. Sites characterized in this way will be candidates for mediating imprinting or male germ cell-specific ICR methylation. (2) Mutate these candidate sites in mice to test their role in mediating imprinting, and (3) identify the proteins binding to these sites. The hypothesis being tested by these aims is that there is a "primary protein difference" between the two germ lines which determines imprinting. These studies are related to health in that genomic imprinting underlies a number of human diseases, e.g., hydatidiform mole, ovarian teratoma, and the Prader-Willi and Beckwith-Wiedemann syndromes. Further understanding of the genetic basis of genomic imprinting will therefore shed light on the genetic basis of these diseases. The research design and methods for achieving this goal will involve (1) in vivo footprinting of the ICR to visualize protein binding sites, (2) targeted mutagenesis of ICR sites in mice and assessment of effects on imprinting by allele-specific expression and methylation analysis of H19 and Igf2, and (3) competitive and supershift electrophoretic mobility shift assays to identify proteins binding to ICR sites. Further assessment of protein identity by examining mouse mutants for candidate proteins.